PIC BASIC TLC5940 code

[Darrel Taylor]

Grayscale values are 12-bits, and you need to send 16 of those values before latching the data, for a total of 192 bits.

The code in the attachment is sending 1,536 bits (16*12*8).
And the method you showed in post #1 would only send 128 (16*8).

And your variable sizes aren't right. fade is a BYTE array, and can't hold 12-bit values. data1 is also a BYTE array, but you're using it like a WORD variable.

I think something like this will work better ...

fade   var WORD[16]
TempW  var WORD

setGreys:
    for z = 0 to 15
        TempW = fade(z)
        shiftout ser, sclk, 1, [TempW\12]
    next z
        
    high xlat
    low xlat  
return

Then you could ...

RampCount  var WORD
ChanCount  var BYTE

RampUP:
    FOR RampCount = 0 to 4088 STEP 8
        FOR ChanCount = 0 to 15
            fade(ChanCount) = RampCount
        NEXT ChanCount
        GOSUB setGreys
        PAUSE 50
    NEXT RampCount
RETURN

RampDN:
    FOR RampCount = 4088 to 0 STEP -8
        FOR ChanCount = 0 to 15
            fade(ChanCount) = RampCount
        NEXT ChanCount
        GOSUB setGreys
        PAUSE 50
    NEXT RampCount
RETURN

Dot correction values are 6-bits. 16 of those would total 96-bits.
<br>

[eletrokat]

Thanks for the help with the variable settings. This definitely helps as I'm sure I'm shifting out the right bits now.

I'm still not getting any response from the TLC5940 though. Now instead of randomly blinking LEDs it's just staying off. I think this is fine because I'm shifting out the right bits, but I'm still confused with the ordering of the LATCH, BLANK, SER.

If you look at the Timing Chart in the spec sheet they show the ordering, but there is this weird thing where they have a separate pulse GSCLK in addition to the main SCLK pulse. I'm not sure why they have this and I don't know how I'm supposed to do both at the same time..

[charudatt]

Take a look at this forum topic , maybe it can help you understand the logic of sending data to the chip.

http://www.picaxeforum.co.uk/showthr...?t=8616&page=3

let us know when you succeed.

[eletrokat]

Hey thanks for that link. That's some really good info. I had to to put this chip on hold because I got swamped with other things, but I was actually going to go back at it this week. This info thread you posted will help.

I'll post my results (or non-results) later on this week.

Thanks.

[wolwil]

How can I get this to stop flickering? It flickers even at 16MHz. With 4 MHz it taks the GSCLK about 250 ms to complete making all the GSDATA that are 0's really stand out causing a flicker.

Is there some other way I could do the GSCLK so I am not just turning the GSCLK pin H/L 4096 times?

ANSEL = 0 'ALL DIGITAL
CMCON = 7 'COMPARATORS OFF
TRISA = %00000000
TRISB = %00000000

DPIN        CON     7
SCLK        CON     6
XLAT        CON     5
BLANK       CON     2
VPRG        var     PORTA.0
DCPRG       var     PORTA.2


C1          var     byte
C2          var     byte
C3          var     word

GSDATA      var     word

GOSUB preset
'***************************************************
'*                    Main Loop                    *
'***************************************************
LOOP:
FOR GSDATA = 4095 TO 0 STEP -16
    gosub setgs
    gosub lightemup
NEXT
FOR GSDATA = 4095 TO 0 STEP 16
    gosub setgs
    gosub lightemup
NEXT
goto LOOP
'***************************************************
'*                      Set GS                     *
'***************************************************
SETGS:
for C1 = 0 to 15                        ' for 16 outputs
    shiftout dpin,sclk,1,[GSDATA\12]    ' shift out 12 bit GS Data
next
    PORTB = %00100000                   ' high xlat
    PORTB = %00000000                   ' low xlat
return
'***************************************************
'*                   LIGHT EM UP                   *
'***************************************************
LIGHTEMUP:
PORTB = %00000100                       ' high blank
PORTB = %00000000                       ' All outputs = ON 
for C3 = 0 to 4095                      ' Generate 4096 GSCLK pulses 
    PORTA = %00001000                   ' GSCLK HIGH 
    PORTA = %00000000                   ' GSCLK LOW 
next
return
'***************************************************
'*                      Preset                     *
'***************************************************
PRESET:
'HIGH DCPRG
HIGH VPRG                               ' Put into dot correction mode
HIGH blank                              ' All outputs OFF
low dpin                                ' Starts dot correction
low xlat
For C1 = 0 to 15                        ' 16 Ports
    shiftout dpin,sclk,1,[%111111\6]    ' 6 bits of Dot Correction per port
next
high xlat                               ' Latch bits into dot correct register
low xlat
low VPRG                                ' Enter GS Mode
for C1 = 0 to 15
    for C2 = 0 to 11
        shiftout dpin,sclk,1,[%111111111111\12] 
    next
next
high xlat                               ' Latch bits into GS register 
low xlat
high SCLK                               '193rd CLK pulse
LOW SCLK
low blank                               ' All outputs = ON 
return

NOTE: This code is basically the same as the above Adruno code but I have a different fade routine. The above code used direct PORT addressing for GSCLK like I did but they were using an 8MHz Adruno for their project and it worked fine. I don't understand why this will not work in PBP with 16 MHz running the show.